Gerardo Roa-Dabike

Postdoctoral Researcher | Machine Learning & Audio Signal Processing | Music & Hearing Technologies

I am a postdoctoral researcher in machine learning and audio signal processing at the University of Sheffield. My research focuses on developing perceptually motivated signal processing and deep learning methods to improve music perception for people with hearing loss. I am a core contributor to the Cadenza challenges, working on dataset design, evaluation methodologies, and benchmark development.

My broader interests include audio representation learning, music information retrieval, and real-time DSP systems, with a particular emphasis on translating research outcomes into robust, real-time solutions for hearing technologies and audio applications.

Not related to my research — just a nice memory from a trip to Chilean Patagonia.

news

Aug 14, 2026	We have good news: the CLIP2 Cadenza Challenge has been accepted as an ICASSP 2027 Signal Processing Grand Challenge. This is the second Cadenza Challenge on Predicting Lyric Intelligibility, launched in June 2026. More details will be available soon on the Cadenza Challenge website.
Jun 01, 2026	Launching my new personal website!

selected publications

Data Paper

The First Cadenza Challenges: Using Machine Learning Competitions to Improve Music for Listeners With a Hearing Loss

Gerardo Roa-Dabike, Michael A. Akeroyd, Scott Bannister, Jon P. Barker, Trevor J. Cox, Bruno Fazenda, Jennifer Firth, Simone Graetzer, Alinka Greasley, Rebecca R. Vos, and William M. Whitmer

IEEE Open Journal of Signal Processing, 2025

DOI Bib Code Website

@article{roadabike2025_ojsp,
  author = {Roa-Dabike, Gerardo and Akeroyd, Michael A. and Bannister, Scott and Barker, Jon P. and Cox, Trevor J. and Fazenda, Bruno and Firth, Jennifer and Graetzer, Simone and Greasley, Alinka and Vos, Rebecca R. and Whitmer, William M.},
  journal = {IEEE Open Journal of Signal Processing},
  title = {The First Cadenza Challenges: Using Machine Learning Competitions to Improve Music for Listeners With a Hearing Loss},
  year = {2025},
  volume = {6},
  number = {},
  pages = {722-734},
  doi = {10.1109/OJSP.2025.3578299},
}

Data Paper

The Cadenza lyric intelligibility prediction (CLIP) dataset

Gerardo Roa-Dabike, Trevor J. Cox, Jon P. Barker, Bruno M. Fazenda, Simone Graetzer, Rebecca R. Vos, Michael A. Akeroyd, Jennifer Firth, William M. Whitmer, Scott Bannister, and Alinka Greasley

Data in Brief, 2026

DOI Bib Code Website

@article{ROADABIKE2026112466,
  title = {The Cadenza lyric intelligibility prediction (CLIP) dataset},
  journal = {Data in Brief},
  volume = {65},
  pages = {112466},
  year = {2026},
  issn = {2352-3409},
  doi = {10.1016/j.dib.2026.112466},
  url = {https://www.sciencedirect.com/science/article/pii/S2352340926000193},
  author = {Roa-Dabike, Gerardo and Cox, Trevor J. and Barker, Jon P. and Fazenda, Bruno M. and Graetzer, Simone and Vos, Rebecca R. and Akeroyd, Michael A. and Firth, Jennifer and Whitmer, William M. and Bannister, Scott and Greasley, Alinka},
}